Pinned Tweet
Fuelio is App of the Day on the App Store in the United States - Apple’s home market 🇺🇸🙀
Thank you Apple, the App Store team, and everyone using @Fuelio
#OnTheAppStore @AppStore #Fuelio #AppStore
It's interesting how Android Studio wasn't mentioned in the keynote. The tooling is changing, and I wonder what the future of Android Studio will look like.
I hope it won't end up being transformed into some Antigravity-Codex-like UI. That's something we already have.
But for sure, it needs to change and adapt to the new world.
What I like - and @matthewmccull started the keynote with this - is that they're not trying to lock you into their own harness and Google's models, like Antigravity does.
Today we're using Claude. In the past, Codex was more popular. In the future, maybe we'll switch to Google models. As we all know, depending on the projects we work on, we need to be able to switch between models and providers.
For some projects, I can't just use a Claude subscription. For private projects, I can play with any model (and that's how I'm working today - trying different models). For enterprise projects, I have to use the models I'm allowed to use.
I like how fast Android DevRel team is moving and in general how #AndroidDev is caching up with all that stuff that is going on in the dev area.
And well, I regret I coudn't make it to Mountain View this year 🥲
#AndroidDevSummit
Yay! 🔥
✨ Gemini Nano 4, our latest on-device AI model, is available to developers through the ML Kit Agentic API.
With Nano 4, you can add an offline AI agent to your Android app.
#AndroidDevSummit
New fear unlocked. At this point I'm not sure I should use Google models on account connected to the rest services at Google.
An update on this Gemini Google policy violation situation.
Thanks to two Googlers (@rodydavis, @_mohansolo) from the @antigravity team who reached out when my post started taking off, we were able to get to the bottom of it.
The 2 violation emails that I got were indeed related to Antigravity and to my attempt the previous day to do defensive research and find security bugs in our own codebase (xitter.gq/ArtemR/status/21145236…).
From what I've gathered, the automated flags are indeed checked by some human team, and that team made a mistake and ended up confirming the flags rather than negating them.
The internal strikes on my account were reversed, so that I wouldn't be penalized for them in the future.
This experience seems to have kicked off an internal effort to improve their processes and prevent similar situations from happening in the future, but I don't know the specifics.
On my part, I've provided constructive feedback about concrete issues that can be improved (the fact that the notifications come after an unspecified amount of time following the activity, don't provide any hints as to what the activity that caused it was, are really vague, and mention unspecified services and content that may get restricted. There's nothing actionable in there, just generic help. It's a really jarring user experience.).
I'd like to thank Rody and Varun for springing into action, resolving this concern, and leading the charge internally to prevent it from happening in the future.
The agentic approach to Android development is huge, and I really like it! #AndroidDevSummit
Can't wait to get my hands on that "prototype" tool @FMuntenescu was showing and check it out!
I wonder how this will work without emulators. Even right now, I created my own workflow with emulators that is working great - but yes, it's sometimes slow a bit (especially for testing).
My Pro sub for Codex ended. I switched to Plus to keep Access to models but right now. Sol, Terra. Nothing is working, same projects, super slow. And usage, I don't think you can even do simple tasks with Codex. Few prompts and it's gone. I'm super confused. I lived with my prv stuff in Codex for a long time. What happened?
Fuelio is “Made for iPhone Duo” from Day 0. We released an app compatible with the new foldable iPhone even before the official premiere. It was a small-medium-sized task, but not that bad, because @Fuelio is a native app.
A few words about this. At first you might think: why do we need to support this? It’s a niche, a $2,000 phone almost no one will get.
Here’s why I think it’s worth it.
Starting in April 2027, whether you like it or not, you’ll have to publish screenshots of your app for iPhone Duo too. But the bigger impact, in my opinion, is that you’re not supporting just Duo. It’s a shared SDK, so many apps will also look better on iPad as a result. Not everything is shared, of course. Tablets don’t support the folding features. But the Duo UI is much closer to a tablet than to a regular phone, so it’s much easier to give the tablet version a better look too. And iPad is not a niche.
For years many developers ignored how their apps look on tablets. I was in that group. I only checked that the app looked OK, and that was all. I didn’t give the iPad any extra work. In most cases the target was iPhone. I think this will change now.
I’m not sure if this was the goal for @Apple and @AppStore Store. If not, it will be a nice side effect.
In my opinion this will change quite fast. All of my apps are ready for Duo from Day 1.
Was it hard? Not so much if you use native stuff. It's getting more problematic if you don't use native. Then you need to adapt Apple's design language and just like Google Apps are using Material both on Android and iOS - it's problematic.
But I think it's a matter of time and don't forget - all apps starting April 2027 will have to put screenshots for Duo in App Store.
Fixes are here but still we need to wait for your distro to put it in official repo.
PHP changes are live. Start patching. php.net/archive/2026.php#202…
Huge thanks to coordinated work by @ThePHPF!
Adrian Kajda retweeted
Remek is currently working on a decision model (i.e. "system one" model, similar to Jev / OpenAI's Decisions API) and he's currently ranked #1 in open weights (#2 overall!) on JevBench, essentially beating almost every commercially released product, including Jev itself that went viral some time ago
what's even crazier is that he's working on it alone, as an after-hours side project, he started a few (two?) weeks ago and he's paying for compute out of his own pocket
basal-omni 🇵🇱 - TOP1 🤩🤩🤩 🌍 na świecie w modelach otwartych oraz TOP2 wśród zamkniętych i otwartych
Wg niezależnego benchmarku modeli system-one gdzie testowane są wszystkjie modele na świecie - JevBench (link do bencha w pierwszym wpisie). Czyli można? Można! Polskie AI 🇵🇱
basal wyprzedził model OpenAI, Microsoft oraz pierwowzór z TypeSafe (jev).
Thank you, @airesearch12 , for the independent evaluation on JevBench.
Average cost for Opus 🙈
📢 Introducing Android Bench 2.0.
We've leveled up our AI evaluation framework to handle real-world, multi-day engineering challenges—from building apps from scratch to migrating cross-platform codebases to Android.
Let’s dive into what’s new. 🧵👇🏽
ALT Android Bench 2.0 dashboard with category filter buttons.
It looks like Antigravity IDE will be ☠️ soon (it's not official info, just guessing based on the updates).
While @antigravity 2.0, CLI, SDK are getting updates quite often. IDE stuck in August.
So it's basically 25% decrease announcement, right?
Starting September 14, we're permanently raising standard weekly limits in Claude Code by 25% for Pro, Max, Team, and seat-based Enterprise plans. Until then, the current 50% increase will be in place.
Readers added context they thought people might want to know
While standard base limits are increasing by 25%, this change represents a net 17% reduction in weekly limits compared to the active 50% promotional limits users currently experience, as confirmed by the official Claude account: x.com/claudedevs/sta…